Maladaptive Exercise in People With a Lifetime History of Eating Disorders: A Multicountry Observational Study

Maladaptive exercise (ME), problematic attitudes or behaviors related to physical activity, is often observed in eating disorders (EDs), but its prevalence, severity, and relationships with ED features remain incompletely characterized. This study aimed to describe lifetime and current ME across ED diagnoses, examine associations with ED symptoms, sociodemographic characteristics, and weight status, and compare ME phenotypes to identify those most strongly linked to ED features.

ME was assessed in 31,671 individuals with lifetime EDs (anorexia nervosa [AN], bulimia nervosa, and binge-eating disorder [BED]) recruited to sites in the United States, Australia, New Zealand, and Sweden as part of the Eating Disorders Genetics Initiative (NCT04378101). We examined one lifetime ME phenotype, capturing any occurrence of ME across the lifespan (derived from the ED100K), and five current ME phenotypes (any driven exercise [Eating Disorder Examination-Questionnaire, EDE-Q], frequency of driven exercise [EDE-Q], regular driven exercise [EDE-Q], compulsive exercise test [CET] total, and dichotomous compulsive exercise [CET]).

Lifetime ME was highly prevalent across countries and ED diagnoses (pooled prevalence 88%). Current ME was also common (40% any driven exercise, 35% compulsive exercise, and 12% regular driven exercise), highest in AN and lowest, but notable, in BED. CET-based phenotypes showed the strongest association with ED symptoms, while EDE-Q-derived regular driven exercise was most strongly associated with underweight (BMI < 18.5). Younger generations exhibited a higher ME risk, likely reflecting societal shifts in body ideals and fitness culture.

This large, multicountry study demonstrates that ME is highly prevalent across EDs, often emerges early, and can persist or recur for decades. Prevalence and severity vary considerably by assessment method. The ED100K, CET, and EDE-Q capture complementary aspects of ME, and incorporating all three may improve clinical care. We recommend that clinicians assess both current and lifetime ME in all patients with EDs. Routine, harmonized assessment will be essential to advancing research and improving patient outcomes.

ClinicalTrials.gov identifier: NCT04378101.